In this chapter, we explored situations of node failures and essential techniques for malfunctioning cluster members, along with some essential cluster concepts in greater depth. In addition to this, we saw how to add cluster resources in order to further configure our newly created cluster into a real-world usage case, which we will deal with during the next chapter.
It is also worth reiterating that there are certain hardware components that we have not been able to discuss in detail, such as fencing devices, and you should take note of the fencing agents and devices (as per pcs stonith list
) and see if any of them applies to the available hardware in your case.
Last but not least, you need to remember that in order to avoid split-brain situations, besides applying thoroughly the concepts outlined in the present chapter, you also need to ensure redundant communication links between the networks where the nodes are located. This will help you prevent a Single Point Of Failure (SPOF...